Previously: binGetSpatialGenes().
BinSpect (Binary Spatial Extraction of genes) is a fast computational method
that identifies genes with a spatially coherent expression pattern.
There are several functions documented together here, mainly differing in
how to provide expression and spatial connectivity/networks information.
When data is in a giotto object, use binSpect() which wraps
binSpectSingle() and binSpectMulti().
binSpect(
gobject,
spat_unit = NULL,
feat_type = NULL,
bin_method = c("kmeans", "rank"),
expression_values = c("normalized", "scaled", "custom"),
subset_feats = NULL,
spatial_network_name = "Delaunay_network",
spatial_network_k = NULL,
reduce_network = FALSE,
kmeans_algo = c("kmeans", "kmeans_arma", "kmeans_arma_subset"),
nstart = 3,
iter_max = 10,
extreme_nr = 50,
sample_nr = 50,
percentage_rank = 30,
do_fisher_test = TRUE,
adjust_method = "fdr",
calc_hub = FALSE,
hub_min_int = 3,
get_av_expr = TRUE,
get_high_expr = TRUE,
implementation = c("data.table", "simple", "matrix"),
group_size = "automatic",
do_parallel = TRUE,
cores = NA,
verbose = TRUE,
knn_params = NULL,
set.seed = deprecated(),
seed = 1234,
bin_matrix = NULL,
summarize = c("p.value", "adj.p.value"),
return_gobject = FALSE
)
binSpectSingleMatrix(
expression_matrix,
spatial_network = NULL,
bin_matrix = NULL,
bin_method = c("kmeans", "rank"),
subset_feats = NULL,
kmeans_algo = c("kmeans", "kmeans_arma", "kmeans_arma_subset"),
nstart = 3,
iter_max = 10,
extreme_nr = 50,
sample_nr = 50,
percentage_rank = 30,
do_fisher_test = TRUE,
adjust_method = "fdr",
calc_hub = FALSE,
hub_min_int = 3,
get_av_expr = TRUE,
get_high_expr = TRUE,
implementation = c("data.table", "simple", "matrix"),
group_size = "automatic",
do_parallel = TRUE,
cores = NA,
verbose = FALSE,
set.seed = deprecated(),
seed = 1234
)
binSpectSingle(
gobject,
spat_unit = NULL,
feat_type = NULL,
bin_method = c("kmeans", "rank"),
expression_values = c("normalized", "scaled", "custom"),
subset_feats = NULL,
spatial_network_name = "Delaunay_network",
reduce_network = FALSE,
kmeans_algo = c("kmeans", "kmeans_arma", "kmeans_arma_subset"),
nstart = 3,
iter_max = 10,
extreme_nr = 50,
sample_nr = 50,
percentage_rank = 30,
do_fisher_test = TRUE,
adjust_method = "fdr",
calc_hub = FALSE,
hub_min_int = 3,
get_av_expr = TRUE,
get_high_expr = TRUE,
implementation = c("data.table", "simple", "matrix"),
group_size = "automatic",
do_parallel = TRUE,
cores = NA,
verbose = TRUE,
set.seed = deprecated(),
seed = 1234,
bin_matrix = NULL
)
binSpectMulti(
gobject,
feat_type = NULL,
spat_unit = NULL,
bin_method = c("kmeans", "rank"),
expression_values = c("normalized", "scaled", "custom"),
subset_feats = NULL,
spatial_network_k = c(5, 10, 20),
reduce_network = FALSE,
kmeans_algo = c("kmeans", "kmeans_arma", "kmeans_arma_subset"),
nstart = 3,
iter_max = 10,
extreme_nr = 50,
sample_nr = 50,
percentage_rank = c(10, 30),
do_fisher_test = TRUE,
adjust_method = "fdr",
calc_hub = FALSE,
hub_min_int = 3,
get_av_expr = TRUE,
get_high_expr = TRUE,
implementation = c("data.table", "simple", "matrix"),
group_size = "automatic",
do_parallel = TRUE,
cores = NA,
verbose = TRUE,
knn_params = NULL,
set.seed = deprecated(),
seed = 1234,
summarize = c("adj.p.value", "p.value")
)giotto object
spatial unit
feature type
method to binarize gene expression
expression values to use
only select a subset of features to test
name of spatial network to use (default = 'spatial_network')
different k's for a spatial kNN to evaluate
default uses the full network
kmeans algorithm to use (kmeans, kmeans_arma, kmeans_arma_subset)
kmeans: nstart parameter
kmeans: iter.max parameter
number of top and bottom cells (see details)
total number of cells to sample (see details)
percentage of top cells for binarization
perform fisher test
p-value adjusted method to use
(see p.adjust)
calculate the number of hub cells
minimum number of cell-cell interactions for a hub cell
calculate the average expression per gene of the high expressing cells
calculate the number of high expressing cells per gene
enrichment implementation (data.table, simple, matrix)
number of genes to process together with data.table implementation (default = automatic)
run calculations in parallel with mclapply
number of cores to use if do_parallel = TRUE
be verbose
list of parameters to create spatial kNN network
deprecated. Use seed param instead
seed for kmeans binarization. When NULL, no seed is set.
Otherwise, accepts a numeric input that will be used as seed.
a binarized matrix, when provided it will skip the binarization process
summarize the p-values or adjusted p-values
whether to return values attached to the gobject or separately (default)
expression matrix
spatial network in data.table format
deprecated, use subset_feats
data.table with results (see details)
We provide two ways to identify spatial genes based on gene expression binarization. Both methods are identicial except for how binarization is performed.
binarize: Each gene is binarized (0 or 1) in each cell with kmeans (k = 2) or based on rank percentile
network: All cells are connected through a spatial network based on the physical coordinates
contingency table: A contingency table is calculated based on all edges of neighboring cells and the binarized expression (0-0, 0-1, 1-0 or 1-1)
For each gene an odds-ratio (OR) and fisher.test (optional) is calculated
Three different kmeans algorithms have been implemented:
kmeans: default, see kmeans
kmeans_arma: from ClusterR, see KMeans_arma
kmeans_arma_subst: from ClusterR, see
KMeans_arma, but randomly subsets the vector
for each gene to increase speed. Change extreme_nr and sample_nr for
control.
Other statistics are provided (optional):
Number of cells with high expression (binary = 1)
Average expression of each gene within high expressing cells
Number of hub cells, these are high expressing cells that have a user defined number of high expressing neighbors
By selecting a subset of likely spatial genes
(e.g. soft thresholding highly variable genes) can accelerate the speed.
The simple implementation is usually faster, but lacks the possibility to
run in parallel and to calculate hub cells.
The data.table implementation might be more appropriate for large datasets
by setting the group_size (number of genes) parameter to divide the
workload.
binSpectSingleMatrix(): binSpect for a single spatial network and a provided
expression matrix
binSpectSingle(): binSpect for a single spatial network
binSpectMulti(): binSpect for multiple spatial kNN networks
g <- GiottoData::loadGiottoMini("visium")
#> 1. read Giotto object
#> 2. read Giotto feature information
#> 3. read Giotto spatial information
#> 3.1 read Giotto spatial shape information
#> cell_spatInfo_spatVector.shp
#> cell
#>
#> 3.2 read Giotto spatial centroid information
#> cell
#>
#> 3.3 read Giotto spatial overlap information
#> No overlaps were found, overlap loading will be
#> skipped
#>
#> 4. read Giotto image information
#> a giotto python environment was found
#> Using python path:
#> "/Users/yuanlab/Library/r-miniconda/envs/giotto_env/bin/pythonw"
binSpect(g)
#>
#> This is the single parameter version of binSpect
#>
#> 1. matrix binarization complete
#>
#> 2. spatial enrichment test completed
#>
#> 3. (optional) average expression of high
#> expressing cells calculated
#>
#> 4. (optional) number of high expressing cells
#> calculated
#> feats p.value estimate adj.p.value score av_expr
#> <char> <num> <num> <num> <num> <num>
#> 1: Shox2 2.079751e-222 24.3725907 3.672044e-220 1.244079e+04 3.980434
#> 2: Ddn 1.845757e-278 19.2760463 1.170210e-275 1.232714e+04 6.032232
#> 3: Hpca 1.055828e-234 12.1576137 3.346976e-232 6.549922e+03 6.930036
#> 4: Zic1 2.316747e-222 12.1139594 3.672044e-220 6.182162e+03 4.279412
#> 5: Cplx2 1.420300e-191 12.5040171 1.286386e-189 5.494801e+03 5.976190
#> ---
#> 630: Vcam1 9.683169e-01 0.9933599 9.744650e-01 3.198207e-02 3.049218
#> 631: Fn1 1.000000e+00 0.9975663 1.000000e+00 0.000000e+00 3.295082
#> 632: Ncf2 1.000000e+00 0.9690611 1.000000e+00 0.000000e+00 3.053523
#> 633: Ndufa4l2 1.000000e+00 0.9933626 1.000000e+00 0.000000e+00 3.195307
#> 634: Vsir 1.000000e+00 0.9988877 1.000000e+00 0.000000e+00 3.032595
#> high_expr
#> <num>
#> 1: 118
#> 2: 428
#> 3: 385
#> 4: 211
#> 5: 459
#> ---
#> 630: 190
#> 631: 193
#> 632: 38
#> 633: 115
#> 634: 177
binSpectSingle(g)
#>
#> This is the single parameter version of binSpect
#>
#> 1. matrix binarization complete
#>
#> 2. spatial enrichment test completed
#>
#> 3. (optional) average expression of high
#> expressing cells calculated
#>
#> 4. (optional) number of high expressing cells
#> calculated
#> feats p.value estimate adj.p.value score av_expr
#> <char> <num> <num> <num> <num> <num>
#> 1: Shox2 2.079751e-222 24.3725907 3.672044e-220 1.244079e+04 3.980434
#> 2: Ddn 1.845757e-278 19.2760463 1.170210e-275 1.232714e+04 6.032232
#> 3: Hpca 1.055828e-234 12.1576137 3.346976e-232 6.549922e+03 6.930036
#> 4: Zic1 2.316747e-222 12.1139594 3.672044e-220 6.182162e+03 4.279412
#> 5: Cplx2 1.420300e-191 12.5040171 1.286386e-189 5.494801e+03 5.976190
#> ---
#> 630: Vcam1 9.683169e-01 0.9933599 9.744650e-01 3.198207e-02 3.049218
#> 631: Fn1 1.000000e+00 0.9975663 1.000000e+00 0.000000e+00 3.295082
#> 632: Ncf2 1.000000e+00 0.9690611 1.000000e+00 0.000000e+00 3.053523
#> 633: Ndufa4l2 1.000000e+00 0.9933626 1.000000e+00 0.000000e+00 3.195307
#> 634: Vsir 1.000000e+00 0.9988877 1.000000e+00 0.000000e+00 3.032595
#> high_expr
#> <num>
#> 1: 118
#> 2: 428
#> 3: 385
#> 4: 211
#> 5: 459
#> ---
#> 630: 190
#> 631: 193
#> 632: 38
#> 633: 115
#> 634: 177
g_expression <- getExpression(g, output = "matrix")
g_spat_net <- getSpatialNetwork(g, output = "networkDT")
binSpectSingleMatrix(
expression_matrix = g_expression,
spatial_network = g_spat_net
)
#> feats p.value estimate adj.p.value score av_expr high_expr
#> <char> <num> <num> <num> <num> <num> <num>
#> 1: 2900040C04Rik 2.824859e-03 Inf 4.033695e-03 Inf 17.00000 2
#> 2: Cbln1 1.788492e-05 Inf 3.048128e-05 Inf 31.33333 3
#> 3: Cd59a 1.788492e-05 Inf 3.048128e-05 Inf 26.33333 3
#> 4: Clic6 2.824859e-03 Inf 4.033695e-03 Inf 16.50000 2
#> 5: Ctxn3 1.788492e-05 Inf 3.048128e-05 Inf 30.00000 3
#> ---
#> 630: Pcolce 1.000000e+00 0 1.000000e+00 0 15.00000 1
#> 631: Prph 1.000000e+00 0 1.000000e+00 0 7.20000 5
#> 632: S100a8 1.000000e+00 0 1.000000e+00 0 8.75000 8
#> 633: Tpm2 1.000000e+00 0 1.000000e+00 0 20.50000 2
#> 634: Vcam1 1.000000e+00 0 1.000000e+00 0 17.00000 1